Low-amplitude, left vagus nerve stimulation significantlyattenuates ventricular dysfunction and infarct size throughprevention of mitochondrial dysfunction during acuteischemia-reperfusion injuryKrekwit Shinlapawittayatorn, MD, PhD,* Kroekkiat Chinda, DVM,* Siripong Palee, PhD,*

Sirirat Surinkaew, PhD,* Kittiya Thunsiri, MEng,* Punate Weerateerangkul, PhD,*

Siriporn Chattipakorn, DDS, PhD,*† Bruce H. KenKnight, PhD,‡ Nipon Chattipakorn, MD, PhD*

From the *Cardiac Electrophysiology Research and Training Center, Department of Physiology, Faculty of Medicine,†Department of Oral Biology and Diagnostic Science, Faculty of Dentistry, Chiang Mai University, Chiang Mai,Thailand and ‡Emerging Therapies, Cyberonics Inc, Houston, Texas.

BACKGROUND Right cervical vagus nerve stimulation (VNS) providescardioprotective effects against acute ischemia-reperfusion injury insmall animals. However, inconsistent findings have been reported.

OBJECTIVE To determine whether low-amplitude, left cervical VNSapplied either intermittently or continuously imparts cardioprotec-tion against acute ischemia-reperfusion injury.

METHODS Thirty-two isoflurane-anesthetized swine (25–30 kg)were randomized into 4 groups: control (sham operated, no VNS),continuous-VNS (C-VNS; 3.5 mA, 20 Hz), intermittent-VNS (I-VNS;continuously recurring cycles of 21-second ON, 30-second OFF), andI-VNS þ atropine (1 mg/kg). Left cervical VNS was appliedimmediately after left anterior descending artery occlusion (60minutes) and continued until the end of reperfusion (120 minutes).The ischemic and nonischemic myocardium was harvested forcardiac mitochondrial function assessment.

RESULTS VNS significantly reduced infarct size, improved ventric-ular function, decreased ventricular fibrillation episodes, andattenuated cardiac mitochondrial reactive oxygen species produc-tion, depolarization, and swelling, compared with the control group.However, I-VNS produced the most profound cardioprotectiveeffects, particularly infarct size reduction and decreased ventricular

This work was supported in part by the Thailand Research Fund andCommission on Higher Education TRF-CHE Research Grant for NewScholar MRG5580125 (to Dr Shinlapawittayatorn), BRG5480003 (to DrChattipakorn), and RTA5580006 (to Dr Chattipakorn) and by Cyberonics.Dr Chattipakorn has served as a member of scientific advisory board forCyberonics. Dr. KenKnight is an employee of Cyberonics. Address reprintrequests and correspondence: Dr Krekwit Shinlapawittayatorn, CardiacElectrophysiology Research and Training Center, Department of Physiol-ogy, Faculty of Medicine, Chiang Mai University, Chiang Mai 50200,Thailand. E-mail address: kshinlap@gmail.com.

1547-5271/$-see front matter B 2013 Heart Rhythm Society. All rights reserved.

fibrillation episodes, compared to both I-VNSþ atropine and C-VNS.These beneficial effects of VNS were abolished by atropine.

CONCLUSIONS During ischemia-reperfusion injury, both C-VNSand I-VNS provide significant cardioprotective effects comparedwith I-VNS þ atropine. These beneficial effects were abolished bymuscarinic blockade, suggesting the importance of muscarinicreceptor modulation during VNS. The protective effects of VNScould be due to its protection of mitochondrial function duringischemia-reperfusion.

KEYWORDS Vagus nerve stimulation; Ischemia-reperfusion injury;Acetylcholine; Heart; Cardioprotection

ABBREVIATIONS AAR ¼ area at risk; C-VNS ¼ continuous-vagusnerve stimulation; ECG ¼ electrocardiographic/electrocardiogram;HR ¼ heart rate; I-VNS ¼ intermittent-vagus nerve stimulation;LAD ¼ left anterior descending; LV ¼ left ventricular; PVC ¼premature ventricular contraction; ROS ¼ reactive oxygen species;VNS ¼ vagus nerve stimulation; VF ¼ ventricular fibrillation; VT ¼ventricular tachycardia

(Heart Rhythm 2013;10:1700–1707) I 2013 Heart Rhythm Society.All rights reserved.

IntroductionAcute myocardial infarction is one of the leading causes ofdeath worldwide.1 Early and successful myocardial reperfusion

with either thrombolytic therapy or primary percutaneouscoronary intervention is the most effective modality forreducing the infarct size and improving clinical outcomes.2

However, the process of abruptly restoring blood flow to theischemic tissue during primary percutaneous coronary inter-vention may result in a devastating cascade of biologicalprocesses, leading to the production of several toxic com-pounds.3,4 This phenomenon, so-called myocardial ischemia-reperfusion injury, can paradoxically reduce the beneficialeffects of myocardial reperfusion.3 Therefore, myocardialischemia-reperfusion injury is considered a major concern inpatients with acute myocardial infarction or those undergoingcoronary artery bypass grafting and transplantation.5,6

http://dx.doi.org/10.1016/j.hrthm.2013.08.009

1701Shinlapawittayatorn et al Vagal Stimulation and Cardiac Mitochondria

Over the past decade, vagus nerve stimulation (VNS)has been shown to exert cardioprotection in both chronicheart failure and ischemic heart diseases.7–12 In ischemichearts, the effects of VNS have been shown to improvecardiac function, limit dispersion of repolarization, pre-vent reperfusion injury, attenuate cardiac remodeling,improve defibrillation efficacy, and decrease the infarctsize.13,14 Despite these well-documented beneficial effectsof VNS, inconsistent findings have been reported. Cur-rently, more than 100,000 patients worldwide havealready been implanted with the left cervical VNS systemfor suppression of epilepsy and depression.15 Left cervicalVNS is preferred over the right because of the greaternumber of cardiac efferent fibers from the right vagusnerve,16 whose stimulation may elicit more frequentundesirable effects. Furthermore, the effect of low-amplitude VNS on cardiac mitochondria has been rarelyinvestigated. Since cardiac mitochondria have been indi-cated as a major determinant of both cardiac cell survivaland arrhythmias, it is important that the effect of leftcervical VNS on cardiac mitochondria during ischemia-reperfusion be explored. Therefore, in the present study,we sought to determine whether left cervical VNS appliedeither intermittently or continuously imparts cardiopro-tection against acute ischemia-reperfusion injury in aswine model.

MethodsAnimal preparationAll experiments were approved by the Institutional AnimalCare and Use Committees of the Faculty of Medicine,Chiang Mai University, Chiang Mai, Thailand. Thirty-twodomestic pigs (25–30 kg) were anesthetized by intramuscu-lar injection of a combination of 4.4 mg/kg of Zoletil (VibbacLaboratories, Carros, France) and 2.2 mg/kg of xylazine(Laboratorios Calier, SA, Barcelona, Spain). After endotra-cheal intubation, anesthesia was maintained by 1.5%–3.0%isoflurane (Abbott Laboratories Ltd, Queenborough, UK)delivered in 100% oxygen. Surface electrocardiogram(ECG) (lead II), femoral arterial blood pressure, heart rate(HR), and rectal temperature were continuously monitored,and all data were recorded for subsequent analysis. Arterialblood gases and electrolytes were also monitored every 30minutes and maintained within acceptable physiologicalranges.17 Under fluoroscopic guidance, platinum-coatedtitanium coil electrodes (34 and 68 mm) were introducedinto and positioned at the right ventricular apex and junctionbetween the right atrium and the superior vena cava,respectively, to deliver electrical shocks when malignantventricular arrhythmias spontaneously occurred duringischemia-reperfusion.17 The chest was opened through a leftthoracotomy. The left anterior descending (LAD) artery wasisolated and occluded by ligature (3-0 silk) 3 cm from the leftmain coronary artery. Complete LAD artery occlusion wasmaintained for 60 minutes and then reperfusion was allowedfor the next 120 minutes.

VNS protocolThe left vagus nerve was surgically isolated (�3 cm, C5–C6level) from the carotid sheath. A VNS lead (Model 304,Cyberonics) with bipolar electrodes (platinum-iridium, 4 mm2

surface area, 6-mm interelectrode spacing) was attached to thevagus nerve using helical fixation elements to assure electrodestability. The cathodic electrode was oriented closest to theheart. The proximal terminal pin of VNS lead was attached toa pulse generator (Demipulse, Model 103, Cyberonics) thatwas used to deliver electrical stimulation during VNS. At thebeginning of the study, the mean PR interval was determinedfrom an average of 10 sinus beats. Then, the 3.5 mAamplitude with 500 ms pulse width and 20 Hz current (basedon the Food and Drug Administration approval for clinicaluse in epilepsy treatment) was set for the left cervical VNS.18

This left cervical VNS was sufficient to produce an increaseof 5–10 ms in PR intervals, and this effect of VNS completelydisappeared as indicated by the return of the PR interval to theprestimulation values within 20 seconds after the cessation ofVNS. Pigs were randomly divided into 4 groups (n ¼ 8 pergroup; Figure 1), and all pigs in each group underwent 60-minute ischemia followed by 120-minute reperfusion. Group1 was the control (sham-operated) group without VNS. Group2 received continuous left cervical VNS (C-VNS; 3.5 mAamplitude, 500 ms pulse duration, and 20 Hz pulse frequency)immediately after LAD artery occlusion and continued untilthe end of reperfusion. Group 3 received intermittent leftcervical VNS (I-VNS; continuous recurring cycles of 21-second ON and 30-second OFF) immediately after LADartery occlusion and continued until the end of reperfusion.Group 4 was similar to group 3 except that atropine (1 mg/kg) was administered intravenously 15 minutes before theleft cervical VNS protocol to block parasympathetic actionson the heart.19 During the ischemia-reperfusion study ineach pig, the cardiac function was continuously monitoredand recorded by using the pressure-volume loop recordingsystem (Model ADV500/ADVantage System, SciSenseInc, London, Canada), as described previously.20 ForECG analysis, the mean baseline PR interval was deter-mined from an average of 10 sinus beats just before LADartery occlusion. The mean PR intervals during the ische-mia and reperfusion periods were analyzed from an averageof 10 consecutive beats before the end of occlusion and theend of reperfusion, respectively.

Evaluation of rhythm disturbancesPremature ventricular contractions (PVCs), ventriculartachycardia (VT), and ventricular fibrillation (VF) weredefined according to the Lambeth convention criteria21 withmore rigorous modifications. Specifically, PVCs weredefined as ventricular contractions without atrial depolariza-tion. VT was defined as more than 6 consecutive PVCs. VFwas characterized by a loss of synchronicity of the ECG plusdecreased amplitude and a precipitous fall in blood pressurefor more than 1 second.

Figure 1 Schematic representation of the study protocols. The ischemic period (60 minutes) was induced by rapid, complete ligation of the LAD coronary artery. Thereperfusion period was 120minutes in duration. In group 1, the pigs were sham operated and then subjected to ischemia (60minutes) and reperfusion (120minutes) withoutVNS. In groups 2–4, VNS was initiated immediately after the LAD artery ligation and continued until the end of reperfusion. In group 4, pigs were given a bolus injectionintravenously of atropine (1 mg/kg) 15 minutes before the initiation of VNS. LAD ¼ left anterior descending; LC ¼ left cervical; VNS ¼ vagus nerve stimulation.

Heart Rhythm, Vol 10, No 11, November 20131702

Infarct size determinationAt the end of each experiment, the heart was removed andirrigated with normal saline to wash out blood fromchambers and vessels. The infarct size was assessed with0.5% Evans Blue and 1.0% triphenyltetrazolium chloridestaining, as described previously.17 The area was measuredby using Image Tool software version 3.0. The mass ratio ofthe area at risk (AAR) to the total ventricular mass and of theinfarct size normalized to the AAR were calculated.

Isolated cardiac mitochondria study protocolCardiac mitochondria were isolated from the ischemic andnonischemic regions by using the technique described pre-viously,22 and the protein concentration was determined byusing the bicinchoninic acid assay. Isolated cardiac mitochon-drial morphology was confirmed by using a transmissionelectron microscope. The measurement of reactive oxygenspecies (ROS) production and mitochondrial membranepotential changes were determined by using a fluorescentmicroplate reader in all groups, as described previously.22 Inshort, the dichlorohydro-fluorescein diacetate dye was used todetermine the level of ROS production in cardiac mitochon-dria. The dichlorohydro-fluorescein diacetate dye could passthrough the mitochondrial membrane and was oxidized todichlorohydro fluorescein by ROS in the mitochondria. The5,5′,6,6′-tetrachloro-1,1′,3,3′-tetraethyl-benzimidazolcarbo-cyanine iodide dye was used to determine the changes in themitochondrial membrane potential. The 5,5′,6,6′-tetrachloro-1,1′,3,3′-tetraethyl-benzimidazolcarbocyanine iodide dye wascharacterized as a cation and remained in the mitochondrialmatrix as a monomer (green fluorescence). However, it couldinteract with anions in the mitochondrial matrix to forman aggregate (red fluorescence). Cardiac mitochondrial

depolarization was indicated by a decrease in the red/greenfluorescence intensity ratio. Moreover, isolated mitochondrialswelling was assessed by measuring changes in the absorb-ance of the suspension wavelength at 540 nm by using amicroplate reader. Cardiac mitochondria (0.4 mg/mL) wereincubated in 2 mL of respiration buffer: 150 mM KCl, 5 mMHEPES, 5 mM K2HPO4 � 3H2O, 2 mM L-glutamate, and 5mM pyruvate sodium salt. Mitochondrial swelling is associ-ated with decreasing absorbance.

Statistical analysisAll data are presented as mean � SD. The normality andequality of variance were tested by using the Shapiro-Wilktest and Levene test, respectively. The mean values between2 groups were compared by using the unpaired Student t test.One-way analysis of variance with Dunnett multiple com-parison test by using the statistical program SPSS17 (SPSS,Inc, Chicago, IL) was used for multiple sets of data. Thelevel of significance for all statistical tests was P o .05.

ResultsEffect of VNS on ECG parameters and left ventricularfunction during the ischemia-reperfusion periodThe electrophysiological effects of VNS were examined in32 pigs in which HR, PR interval, and left ventricular (LV)function were measured continuously during the ischemia-reperfusion period. In the control group, the HR during theischemic period increased significantly when compared withthe baseline (Figure 2A). Interestingly, HRs at baseline andduring ischemia and reperfusion periods were not different inVNS-treated groups (n¼ 8 per group), either in the presenceor absence of muscarinic blockade by atropine, indicatingthat the effects of VNS prevented increased HR during

I-VN

Sþ

atropine

Reperfusion

Baselin

eIschem

iaReperfusion

19�

724

�3

10�

3*10

�1*

42�

754

�4

27�

7*29

�8*

72�

583

�4

74�

379

�3

16�

212

�2

21�

2*22

�2*

Values

arepresentedas

mean�

SD.

rmittent-vagus

nervestimulation;

SV¼

stroke

volume;

VNS¼

vagus

1703Shinlapawittayatorn et al Vagal Stimulation and Cardiac Mitochondria

ischemia-reperfusion injury independent of muscarinicreceptor activation. Nevertheless, while statistically there isa difference in HR (Figure 2A), it is possible that physio-logically there appears to be no difference between groupsand conditions. The PR interval during the baseline, ische-mic, and reperfusion periods was not different in control,C-VNS, and I-VNS groups (Figure 2B). However, PRintervals during ischemic and reperfusion periods weresignificantly reduced in the I-VNSþatropine group, suggest-ing that sympathetic activity might be relatively increased bymuscarinic blockade. The effect of VNS on LV functionalperformance is shown in Table 1. In the control group, thestroke volume and the ejection fraction were significantlydecreased during the ischemia and reperfusion periodscompared with the baseline. The end-diastolic pressure wassignificantly increased during the ischemia and reperfusionperiods compared with the baseline. C-VNS attenuated thedecrease in stroke volume and ejection fraction only during

Figure 2 Effect of VNS on heart rate and PR interval during the ischemia-reperfusion period. In the control group, the heart rate during the ischemic periodwas significantly increased when compared with the baseline period.A:The heartrate during baseline and during ischemia and reperfusion periods were notdifferent in VNS-treated groups (n¼ 8 per group), indicating that VNS preventedthe increase in the heart rate during the ischemic period independent of muscarinicreceptor activation. B: Mean PR interval during the baseline, ischemic, andreperfusion periods was not different in control, C-VNS, and I-VNS groups.However, PR intervals during ischemic and reperfusion periods were significantlyreduced in the I-VNSþatropine group. Data are presented as mean� SD. *Po.05 vs baseline. C-VNS ¼ continuous-vagus nerve stimulation; I-VNS ¼intermittent-vagus nerve stimulation; VNS ¼ vagus nerve stimulation. Ta

ble1

Effectsof

VNSandmuscarin

icreceptor

inhibition

onpressure-volum

eloop

–deriv

edfunction

alparameters

Control

C-VN

SI-VN

S

Parameter

Baselin

eIschem

iaReperfusion

Baselin

eIschem

iaReperfusion

Baselin

eIschem

ia

SV(m

L)24

�1

8�

2*8�

2*23

�2

10�

2*16

�6

22�

215

�5

EF(%

)55

�6

26�

6*26

�4*

46�

428

�2*

35�

852

�5

38�

6ESP(m

mHg)

85�

977

�6

87�

684

�4

75�

482

�3

76�

669

�3

EDP(m

mHg)

14�

422

�2*

24�

5*17

�8

23�

6*25

�5*

13�

315

�3

Summaryof

leftventricular

(LV)

function

alparametersat

baselin

e,at

theendof

ischem

ia,andat

theendof

reperfusion(n

¼8pergroup).

C-VN

S¼

continuous-vagus

nervestimulation;

EDP¼

end-diastolic

pressure;E

F¼

ejection

fraction

;ESP

¼end-systolicpressure;I-VNS

¼inte

nervestimulation.

* Po

.05vs

baselin

e.

Heart Rhythm, Vol 10, No 11, November 20131704

the reperfusion period. Interestingly, I-VNS preserved LVfunctional performance during the ischemia and reperfusionperiods. However, the beneficial effect of VNS was com-pletely abolished by the administration of atropine.

VNS and the occurrence of cardiac arrhythmia duringischemia-reperfusion periodFigure 3A shows examples of ECG tracings at baseline andafter LAD artery occlusion. PVCs were markedly decreasedfor both C-VNS and I-VNS groups during ischemic andreperfusion periods compared with the control group(Figure 3B). This effect was significantly reduced byatropine, indicating that VNS reduces the frequency ofspontaneous PVCs during the ischemic and reperfusionperiods through a pathway that at least partially involvesmuscarinic receptor activation. Moreover, the number of VT/VF episodes was significantly decreased in the I-VNS groupduring the reperfusion period whereas the number of VT/VF

Figure 3 VNS and the occurrence of cardiac arrhythmia during ischemia-repedescending artery occlusion. B: PVCs were markedly decreased in both C-VNS anthat in the control group. This effect was attenuated by atropine, indicating that Vperiods through muscarinic receptor activation. C: The number of VT/VF episodeswhereas the number of VT/VF episodes increased significantly in the I-VNSþ atroD: Time to VT/VF episode onset was not significantly different between groups d*P o .05 vs control; †P o .05 vs I-VNS. C-VNS ¼ continuous-vagus nerve sstimulation; PVC ¼ premature ventricular contraction; VF ¼ ventricular fibrillatio

was increased significantly in the I-VNS þ atropine groupduring ischemic and reperfusion periods compared with theI-VNS group (Figure 3C). However, time to VT/VF onsetwas not significantly different among groups during bothischemic and reperfusion periods (Figure 3D). In the presentstudy, PVCs did not return to control values with atropinewhereas VT/VF did return to control values, suggesting thatsome other factors might suppress PVCs during reperfusionwith VNS and that PVCs might not be related to VT/VFoccurrence during reperfusion with VNS.

Myocardial infarct sizeMyocardial infarct size was expressed as the percentage ofthe AAR. The AAR, expressed as a percentage of the totalventricular mass, was not different among groups (control:32.8% � 4.9%; C-VNS: 37.1% � 2.0%; I-VNS: 35.6% �5.8%; I-VNSþatropine: 38.1%� 4.7%; P¼ not significant).C-VNS significantly reduced mean infarct size by 60%,

rfusion injury. A: The ECG (lead II) recorded before and after left anteriord I-VNS groups during the ischemic and reperfusion periods, compared withNS prevented the formation of PVCs during the ischemic and reperfusionwas significantly reduced in the I-VNS group during the reperfusion period,pine group during ischemic and reperfusion periods than in the I-VNS group.uring ischemic and reperfusion periods. Data are presented as mean � SD.timulation; ECG ¼ electrocardiogram; I-VNS ¼ intermittent-vagus nerven; VNS ¼ vagus nerve stimulation; VT ¼ ventricular tachycardia.

1705Shinlapawittayatorn et al Vagal Stimulation and Cardiac Mitochondria

whereas and I-VNS significantly reduced mean infarct sizeby 89%. The mean infarct size reduction observed for I-VNSwas 48% more than that observed for C-VNS. The admin-istration of atropine totally abolished the cardioprotectiveeffects of VNS with respect to infarct size (Figure 4).

Effect of VNS on cardiac mitochondria afterthe ischemia-reperfusion periodCardiac mitochondrial dysfunction including increased ROSproduction, mitochondrial membrane depolarization, and mito-chondrial swelling has been shown to participate in myocytedysfunction and degradation of cardiac contractile performance,arrhythmias, and myocyte apoptosis and infarct size during andafter ischemia-reperfusion injury. Our results demonstrated thatboth C-VNS and I-VNS decreased cardiac mitochondrial ROSproduction (Figure 5A) and prevented cardiac mitochondrialmembrane depolarization (Figure 5B) and cardiac mitochon-drial swelling (Figure 5C) in the ischemic myocardium,compared with I-VNS þ atropine. Again, these beneficialeffects on cardiac mitochondria were abolished by atropine,suggesting the importance of muscarinic receptor activationduring VNS. However, mitochondrial ROS production(Figure 5A) and mitochondrial membrane depolarization

Figure 4 Determination of myocardial infarct size. Myocardial infarctsize was expressed as the percentage of AAR. Regional myocardial bloodflow during ischemia and the AAR, expressed as a percentage of the totalventricular mass, were not different between groups (control: 32.8% �4.9%; C-VNS: 37.1% � 2.0%; I-VNS: 35.6% � 5.8%; I-VNS þ atropine:38.1% � 4.7%; P¼NS). Interestingly, VNS significantly reduced myocar-dial infarct size and this effect was reversed by atropine (control: 45.7% �14.2%, n¼ 7; C-VNS: 18.5%� 10.5%, n¼ 8; I-VNS: 5.1%� 3.1%, n¼ 6;I-VNS þ atropine: 52.9% � 8.7%, n ¼ 6). Representative pictures afterEvans Blue (Sigma-Aldrich) and triphenyltetrazolium chloride staining areshown in the inset. Blue indicates nonthreatened myocardium; red indicatesthe noninfarcted area within the AAR; and white indicates myocardialinfarction. Data are presented as mean� SD. *Po .05 vs control; †Po .05vs C-VNS; ‡P o .05 vs I-VNS. AAR ¼ area at risk; C-VNS ¼ continuous-vagus nerve stimulation; I-VNS ¼ intermittent-vagus nerve stimulation;NS ¼ not significant; VNS ¼ vagus nerve stimulation.

(Figure 5B) were partially reversed by atropine in the VNS-treated groups, suggesting the presence of an additional non-muscarinic modulation. Electron photomicrographs demon-strated that in the ischemic area of the control group,ischemia-reperfusion induced severe mitochondrial damage,as indicated by the loss of cristae (Figure 6). Interestingly, thecardiac mitochondria in the ischemic myocardium in both C-VNS and I-VNS groups were significantly preserved afterischemia-reperfusion period, and this effect was not presentafter the administration of atropine (Figure 6).

DiscussionTo our knowledge, this is the first demonstration that I-VNSprovides more robust efficacy than C-VNS with respect to theprevention of cardiac mitochondrial dysfunction during theischemia-reperfusion period. Moreover, our finding confirmsprevious reports,9,14,23,24 which also indicated that VNSexerted cardioprotection and improved LV function duringischemia-reperfusion period. Our findings on the betterefficacy of I-VNS than C-VNS suggest that intermittentVNS may prevent adaptation of neural structures involved incardioprotection, sometimes referred to as “fade.”25 In thepresent study, our finding that VNS significantly decreased theoccurrence of PVCs and the number of spontaneous VT/VFepisodes is consistent with previous studies that VNS exertedthe antiarrhythmic effects during the ischemia-reperfusionperiod. Waxman et al26 provided the first evidence that someVT could be modulated by vagal activation and that ventricularautomaticity was reduced by enhanced vagal tone. Moreover,an elegant study by Schwartz and colleagues9 in consciousdogs clearly demonstrated that enhanced vagus nerve activityby means of right cervical VNS prevented spontaneousventricular tachyarrhythmias in a model with healed myocardialinfarction, exercise testing, and intermittent ischemia. In thepresent study, left cervical VNS exerted cardioprotective effectswithout HR alteration, supporting that the benefits of VNScould be independent of HR changes. Recently, left-sided low-level VNS has been shown to suppress stellate ganglion nerveactivity and reduce the incidences of paroxysmal atrial tachyar-rhythmias in ambulatory dogs.27 In addition, left-sided low-level VNS has been shown to upregulate small conductancecalcium-activated potassium channels in the stellate ganglion.28

These changes might be the mechanistic insight underlying theantiarrhythmic effect of low-level VNS.

In the present study, our finding that VNS preserved LVfunctional performance is consistent with previous studies. Inrats subjected to global ischemia-reperfusion with intact vagalinnervation, right cervical VNS-treated LV showed significantlybetter performance throughout the 120-minute reperfusionperiod and that VNS exerted a marked anti-infarct effectirrespective of the HR compared with sham stimulation.29 Thus,the preserved LV function provided by VNS might exertthrough a reduction in the myocardial infarct size. In the presentstudy, both C-VNS and I-VNS significantly reduced myocardialinfarct size by 60% and 89%, respectively. This finding isconsistent with previous both in vitro and in vivo studies

Figure 5 Effect of VNS on cardiac mitochondria after ischemia and reperfusion periods. Both C-VNS and I-VNS decreased mitochondrial ROS production(A) and prevented mitochondrial membrane depolarization (B) and mitochondrial swelling (C) in the ischemic myocardium, compared with I-VNS þ atropine.Again, these effects were diminished by atropine, suggesting the importance of muscarinic modulation during VNS. However, mitochondrial ROS production(panel A) and mitochondrial membrane depolarization (panel B) were only partially attenuated by atropine in the VNS-treated groups, suggesting the presence ofadditional nonmuscarinic modulation. Data are presented as mean � SD. *P o .05 vs control; †P o .05 vs I-VNS. C-VNS ¼ continuous-vagus nervestimulation; I-VNS ¼ intermittent-vagus nerve stimulation; ROS ¼ reactive oxygen species; VNS ¼ vagus nerve stimulation.

Heart Rhythm, Vol 10, No 11, November 20131706

showing that increased vagus nerve activity mainly via rightcervical VNS reduced the ratio of infarct size to the AAR,suggesting that VNS has an anti-apoptotic effect on themyocardium during and after the ischemia-reperfusion period.Kakinuma et al23 reported that infarct size was significantlyreduced by 25% in right cervical VNS-treated rat hearts.Moreover, in a global ischemia-reperfusion rat model (30minutes of ischemia and 120 minutes of reperfusion) with anintact vagal innervation, right-sided VNS has been shown toexert a marked anti-infarct effect irrespective of the HR.29

Furthermore, early VNS has been shown to reduce infarct sizeand ventricular remodeling by inhibiting the increase inmyocardium interstitial myoglobin,30 norepinephrine,31 andmatrix metalloproteinase levels32 associated with myocardialischemia-reperfusion injury. Recently, it has been shown that the“efferent VNS” could inhibit the myocardial ischemia-reperfusion injury mainly through its nicotinic inhibition.14

Although we did not determine the effects of right cervicalVNS, a large number of experimental studies have reported thecardioprotective effects of right cervical VNS, including infarctsize and arrhythmia reduction.33,34 Therefore, right-sided VNScould have provided similar benefits as observed in the presentstudy. However, left-sided VNS would be less likely to interferewith the HR.

In the present study, we have shown that one potentialpossible mechanism of the pronounced cardioprotectiveeffect of VNS during ischemia-reperfusion could be due toits effects on the cardiac mitochondria. Increased ROS

production and oscillation of the mitochondrial membranepotential have been shown to play a crucial role in thegenesis of cardiac arrhythmias and myocardial infarction.35

Our data indicate that mitochondrial membrane potential isstabilized by VNS, thereby protecting mitochondria fromaberrant changes in membrane potential. This could lead tothe reduction of arrhythmias occurred during ischemia-reperfusion in the VNS group. Moreover, mitochondrialROS reduction and decreased mitochondrial swelling couldbe responsible for decreased infarct size in the VNS group.These beneficial effects of VNS were abolished by atropine,suggesting a dominant muscarinic receptor involvement.

Study limitationsThere are several limitations in the present study. Our studywas conducted in anesthetized healthy animals, whereas mostvictims of cardiac arrest have significant coronary lesions.Furthermore, VNS treatment was conducted immediately afterLAD artery occlusion. The time schedule of treatment in thisexperimental study may differ from the clinical condition.

ConclusionsWe have thus provided compelling evidence that during acuteischemia-reperfusion period, both C-VNS and I-VNS providesignificant cardioprotection and improve LV function in a largeanimal model of ischemia-reperfusion injury. However, I-VNSprovides more robust efficacy than C-VNS with respect to

Figure 6 Representative electron photomicrographs of a cardiac mito-chondrial ultrastructure. In ischemic area of the control group, ischemia-reperfusion–induced severe mitochondrial damage was observed. Note themitochondrial swelling accompanied by disruption in membrane integrity.However, both C-VNS and I-VNS significantly protected cardiac mitochon-drial swelling after ischemia-reperfusion injury and this effect was abolishedby atropine. C-VNS ¼ continuous-vagus nerve stimulation; I-VNS ¼intermittent-vagus nerve stimulation; VNS ¼ vagus nerve stimulation.

1707Shinlapawittayatorn et al Vagal Stimulation and Cardiac Mitochondria

infarct size reduction and reperfusion arrhythmia prevention. Apotential mechanism of such cardioprotection of VNS isassociated with its prevention of cardiac mitochondrial dys-function during ischemia-reperfusion. Thus, these data stronglysupport the notion that VNS is emerging as a promisingtherapeutic modality in combination with reperfusion therapyfor protecting myocardium at risk of ischemia-reperfusioninjury due to coronary artery disease.

References1. Yasuda S, Shimokawa H. Acute myocardial infarction: the enduring challenge for

cardiac protection and survival. Circulation 2009;73:2000–2008.2. Yellon DM, Hausenloy DJ. Myocardial reperfusion injury. N Engl J Med

2007;357:1121–1135.3. Entman ML, Smith CW. Postreperfusion inflammation: a model for reaction to

injury in cardiovascular disease. Cardiovasc Res 1994;28:1301–1311.4. Hawkins HK, Entman ML, Zhu JY, et al. Acute inflammatory reaction after

myocardial ischemic injury and reperfusion: development and use of aneutrophil-specific antibody. Am J Pathol 1996;148:1957–1969.

5. Ishii H, Ichimiya S, Kanashiro M, et al. Impact of a single intravenousadministration of nicorandil before reperfusion in patients with ST-segment-elevation myocardial infarction. Circulation 2005;112:1284–1288.

6. Wu ZK, Laurikka J, Saraste A, et al. Cardiomyocyte apoptosis and ischemicpreconditioning in open heart operations. Ann Thorac Surg 2003;76:528–534.

7. Engelstein ED. Prevention and management of chronic heart failure withelectrical therapy. Am J Cardiol 2003;91:62F–73F.

8. Li M, Zheng C, Sato T, Kawada T, Sugimachi M, Sunagawa K. Vagal nervestimulation markedly improves long-term survival after chronic heart failure inrats. Circulation 2004;109:120–124.

9. Vanoli E, De Ferrari GM, Stramba-Badiale M, Hull SS Jr, Foreman RD,Schwartz PJ. Vagal stimulation and prevention of sudden death in consciousdogs with a healed myocardial infarction. Circ Res 1991;68:1471–1481.

10. De Ferrari GM, Crijns HJ, Borggrefe M, et al. Chronic vagus nerve stimulation: anew and promising therapeutic approach for chronic heart failure. Eur Heart J2011;32:847–855.

11. Schwartz PJ, De Ferrari GM, Sanzo A, et al. Long term vagal stimulation inpatients with advanced heart failure: first experience in man. Eur J Heart Fail2008;10:884–891.

12. Hauptman PJ, Schwartz PJ, GoldMR, et al. Rationale and study design of the increaseof vagal tone in heart failure study: INOVATE-HF. Am Heart J 2012;163:954–962.

13. Zhao M, Sun L, Liu JJ, Wang H, Miao Y, Zang W. Vagal nerve modulation: apromising new therapeutic approach for cardiovascular diseases. Clin ExpPharmacol Physiol 2012;39:701–705.

14. Calvillo L, Vanoli E, Andreoli E, et al. Vagal stimulation, through its nicotinicaction, limits infarct size and the inflammatory response to myocardial ischemiaand reperfusion. J Cardiovasc Pharmacol 2011;58:500–507.

15. Terry R. Vagus nerve stimulation: a proven therapy for treatment of epilepsy strives toimprove efficacy and expand applications. In: Engineering in Medicine and BiologySociety, 2009. Annual International Conference of the IEEE;2009:4631–4634.

16. Saper CB, Kibbe MR, Hurley KM, et al. Brain natriuretic peptide-likeimmunoreactive innervation of the cardiovascular and cerebrovascular systemsin the rat. Circ Res 1990;67:1345–1354.

17. Chinda K, Palee S, Surinkaew S, Phornphutkul M, Chattipakorn S, ChattipakornN. Cardioprotective effect of dipeptidyl peptidase-4 inhibitor during ischemia-reperfusion injury. Int J Cardiol 2013;31:451–457.

18. Groves DA, Brown VJ. Vagal nerve stimulation: a review of its applications andpotential mechanisms that mediate its clinical effects. Neurosci Biobehav Rev2005;29:493–500.

19. Charlier R. Cardiac actions in the dog of a new antagonist of adrenergic excitationwhich does not produce competitive blockade of adrenoceptors. Br J Pharmacol1970;39:668–674.

20. Lewis ME, Al-Khalidi AH, Bonser RS, et al. Vagus nerve stimulation decreasesleft ventricular contractility in vivo in the human and pig heart. J Physiol2001;534:547–552.

21. Walker MJ, Curtis MJ, Hearse DJ, et al. The Lambeth conventions: guidelines forthe study of arrhythmias in ischaemia infarction, and reperfusion. Cardiovasc Res1988;22:447–455.

22. Thummasorn S, Kumfu S, Chattipakorn S, Chattipakorn N. Granulocyte-colonystimulating factor attenuates mitochondrial dysfunction induced by oxidativestress in cardiac mitochondria. Mitochondrion 2011;11:457–466.

23. Kakinuma Y, Ando M, Kuwabara M, et al. Acetylcholine from vagal stimulationprotects cardiomyocytes against ischemia and hypoxia involving additive non-hypoxic induction of HIF-1alpha. FEBS Lett 2005;579:2111–2118.

24. Sabbah HN, Ilsar I, Zaretsky A, Rastogi S, Wang M, Gupta RC. Vagus nervestimulation in experimental heart failure. Heart Fail Rev 2011;16:171–178.

25. Martin P, Levy MN, Matsuda Y. Fade of cardiac responses during tonic vagalstimulation. Am J Physiol 1982;243:H219–H225.

26. Waxman MB, Sharma AD, Asta J, Cameron DA, Wald RW. The protective effectof vagus nerve stimulation on catecholamine-halothane-induced ventricularfibrillation in dogs. Can J Physiol Pharmacol 1989;67:801–809.

27. Shen MJ, Shinohara T, Park HW, et al. Continuous low-level vagus nervestimulation reduces stellate ganglion nerve activity and paroxysmal atrialtachyarrhythmias in ambulatory canines. Circulation 2011;123:2204–2212.

28. Shen MJ, Hao-Che C, Park HW, et al. Low-level vagus nerve stimulationupregulates small conductance calcium-activated potassium channels in thestellate ganglion. Heart Rhythm 2013;10:910–915.

29. Katare RG,AndoM,KakinumaY, et al. Vagal nerve stimulation prevents reperfusioninjury through inhibition of opening of mitochondrial permeability transition poreindependent of the bradycardiac effect. J Thorac Cardiovasc Surg 2009;137:223–231.

30. Kawada T, Yamazaki T, Akiyama T, et al. Vagal stimulation suppresses ischemia-induced myocardial interstitial myoglobin release. Life Sci 2008;83:490–495.

31. Kawada T, Yamazaki T, Akiyama T, et al. Vagal stimulation suppresses ischemia-induced myocardial interstitial norepinephrine release. Life Sci 2006;78:882–887.

32. Uemura K, Li M, Tsutsumi T, et al. Efferent vagal nerve stimulation inducestissue inhibitor of metalloproteinase-1 in myocardial ischemia-reperfusion injuryin rabbit. Am J Physiol Heart Circ Physiol 2007;293:H2254–H2261.

33. Ando M, Katare RG, Kakinuma Y, et al. Efferent vagal nerve stimulation protectsheart against ischemia-induced arrhythmias by preserving connexin43 protein.Circulation 2005;112:164–170.

34. Uemura K, Zheng C, Li M, Kawada T, Sugimachi M. Early short-term vagalnerve stimulation attenuates cardiac remodeling after reperfused myocardialinfarction. J Card Fail 2010;16:689–699.

35. Brown DA, O’Rourke B. Cardiac mitochondria and arrhythmias. Cardiovasc Res2010;88:241–249.